Imagine you're a business owner and you are about to launch an ad campaign for a new product. You currently have data on your customers lifetime purchases
and purchases in the past year.
In order to make the best use of your advertising budget you want to know which of your customers are more likely to buy this new product.
To do this, you first need to group them into categories... But where does each category start or end?
Instead of trying to answer that question yourself- allow the computer to find optimal groups!
Select how many groups you'd like and watch as the algorithm works by grouping then resetting the group centers!
Select K:
K-means is an unsupervised Clustering Algorithm and
it is especially useful for categorizing large multi-variable datasets. Of course, in real-world usage the algorithm would Group and reset repeatedly until it found a sufficient answer. This grouping can be enhanced by using the average and standard deviations for the groups in order to fit different "shapes" of data groupings.